Milvus is an open-source vector database built for billion-scale similarity search. It powers production AI applications at companies like Salesforce, PayPal, and Shopee.
What Is Milvus?
Milvus is purpose-built for vector similarity search with support for multiple index types, hybrid search, and multi-tenancy.
Zilliz Cloud free tier:
- 2 collections
- 1M vectors
- No credit card
Quick Start
# Milvus Lite (embedded, no Docker)
pip install pymilvus
Python SDK
from pymilvus import MilvusClient
client = MilvusClient("milvus_demo.db") # Embedded mode
# Create collection
client.create_collection(collection_name="docs", dimension=384)
# Insert vectors
data = [
{"id": 1, "vector": [0.1]*384, "text": "AI guide"},
{"id": 2, "vector": [0.2]*384, "text": "Python tutorial"}
]
client.insert(collection_name="docs", data=data)
# Search
results = client.search(
collection_name="docs",
data=[[0.1]*384],
limit=5,
output_fields=["text"]
)
print(results[0])
REST API
# Search vectors
curl -X POST http://localhost:9091/api/v1/search \
-d '{"collection_name":"docs","vectors":[[0.1,0.2,...]],"limit":5}'
# Insert
curl -X POST http://localhost:9091/api/v1/entities \
-d '{"collection_name":"docs","fields_data":[...]}'
Use Cases
- RAG — retrieval for LLM apps
- Image search — visual similarity
- Recommendations — content/product matching
- Anomaly detection — find outliers
- Drug discovery — molecular similarity
Need web data at scale? Check out my scraping tools on Apify or email spinov001@gmail.com for custom solutions.
Top comments (0)